/robowaifu/ - DIY Robot Wives

Advancing robotics to a point where anime catgrill meidos in tiny miniskirts are a reality.

LynxChan updated to 2.5.7, let me know whether there are any issues (admin at j dot w).

Reports of my death have been greatly overestimiste.

Still trying to get done with some IRL work, but should be able to update some stuff soon.


Max message length: 6144

Drag files to upload or
click here to select them

Maximum 5 files / Maximum size: 20.00 MB


(used to delete files and postings)

General Robotics/A.I. news and commentary Robowaifu Technician 09/18/2019 (Wed) 11:20:15 No.404
Anything in general related to the Robotics or A.I. industries, or any social or economic issues surrounding it (especially of RoboWaifus). www.therobotreport.com/news/lets-hope-trump-does-what-he-says-regarding-robots-and-robotics https://archive.is/u5Msf blogmaverick.com/2016/12/18/dear-mr-president-my-suggestion-for-infrastructure-spending/ https://archive.is/l82dZ >=== -add A.I. to thread topic
Edited last time by Chobitsu on 12/17/2020 (Thu) 20:16:50.
>>6868 The AI race is pretty much a farce now. The only funding that goes towards AI is deep learning models requiring giant arrays of GPUs and zero towards theoretical understanding. Basically corporations are just milking governments for money. Very few people in the research community have any clue what they're doing and ironically most of the advances come from brute force algorithms trying different architectures, which speaks volumes about the quality of research going on. They're not really outperforming chance beyond a few dozen talented researchers. The lefties are also nerfing and banning research so there's no worry about them getting ahead. A lot of researchers have gotten fed up with the politics and quit academia to pursue business or their own independent research. There's a huge growing shortage of AI engineers right now. Businesses are desperate for people who understand AI and willing to pay six figures but hundreds of thousands of jobs go unfilled every month. Many don't even care if you have a degree or not so long as you're self-motivated, self-disciplined and know what you're doing. The whole world is incompetent in AI, especially China, except they're masters in bullshiting. To an investor Chinese AI companies look and sound good but really their research papers are just trash that make small improvements to other people's innovations while not understanding why their improvements work. Unfortunately in the West people have taken the bait that AI is a meme so no one even tries at it, meanwhile Chinese AI startups are being flooded with investor money. If anything, it's not an intellectual race but a money race. Despite the shitshow going on, at the rate AI is progressing half of people will be out of jobs in the next five years, not including losses to government lockdowns and house arrest, which why the financial collapse is inevitable. It's just a matter of time before the relief money runs out, probably sometime around Q3/Q4 2021, and all these people default on their debts or we go into hyperinflation, unless the lockdowns are stopped and the millions of small businesses lost are somehow resurrected. Some developing countries are already beginning to default on their debts. Banks are forecasting mortgage defaults to skyrocket next year. I'm speculating governments will allow people to stay in their homes but they will no longer own them, allowing them to be kicked out and moved around at any time like they do in China. In places like Canada they're building barb-wire concentration camps, which according to their own documents could be used to shelter homeless people. I imagine they will sell it to people as helping them in a crisis (one that they created, as philanthropists do) and people will have no idea they're being rounded up into gulags, while the rich continue to own everything and enjoy their lives. This will channel people's anger towards a real communist revolution, unless they wake up to what's going on. Don't count on that though unless people create advanced AI systems and have the necessary infrastructure to reach out to hundreds of millions of people and breakthrough their conditioning or manage to create millions of small businesses and jobs for people. Most people are happy they lost their jobs and get to live off 'free' goodies. They won't realize the grave mistake they made until they try to return to work and find out AI is doing their job and they have to pay the piper. I honestly think it's too late to change the course of things now but it's not the end of the world. People just have to survive the crash and keep moving forward. With all this extra cash in people's hands it's an excellent time to make money. The best path forward I can see is to become self-sufficient off-the-grid and then support others to become self-sufficient, and those who can't do that on their own will need to make friends with people who can.
>>6884 Thanks again, Chobitsu! Will definitely make use of that. >>6890 >relief money runs out, probably sometime around Q3/Q4 2021, and all these people default on their debts >Communist Revolution This is what scares me the most. There are BIG riots coming at some point after this pandemic anon. When people realise that things are not going to get much better. Human relationships are fickle, weak things at the best of times. So many divorces and lots of domestic violence going on (mainly due to poverty). And these people have no loyal robowaifu for support! Nowadays, normie society is so hostile to everyone caught up in it. Most of them are just wearing masks and the strain of maintaining their façade of fake happiness and optimism is obvious. I prefer to avoid that nightmare entirely and hide away to research, design and tinker on my robowaifu. It's the only way I stay sane. Plus eventually if we work hard enough we may have something to release to normie society that will ease some of their suffering. The normies may call us names for having relationships with "objects", but I've read how eager men are to engage with realistic looking sex-dolls https://www.zdnet.com/article/sex-robot-molested-destroyed-at-electronics-show/ And THAT was in the middle of an electronics exhibition. Imagine how they'll be when no-one is looking XD. Just watch what happens when even a semi-functional, friendly robowaifu becomes available! The so-called normies are all desperate for a dose of immortal, synthetic affection anon. It's the cure they don't know/cannot admit that they need. They may mock us now, but in the future they'll be grateful.
Open file (287.40 KB 960x670 waifu_mother.png)
>>6904 >"...which have uniformly unrealistic physical characteristics," kek. >WAHHH! They're too perfect! This isn't fair
>>6904 Normies are pretty vicious. They treat women the same way, as much as they can get away with at least. We're really just one food shortage away from people killing each other in the streets. I'm astonished when I go to the city and see people raging at each other and hating living there so much but they're so accustomed to it they don't even realize how miserable they are. I live in middle of nowhere and don't talk to my neighbors much but we're all friends out here and help each other out whenever necessary. If one of them went hostile in a crisis they'd get teamed on and taken out. In a city though it would be a free-for-all deathmatch. That's just the nature of human relationships. No matter how much good you've done if you do one wrong thing too far, you're gone, and in the city there is no cohesion of values. People use to laugh at using the internet too and say all kinds of stupid shit about it but now they all use it 4+ hours a day. I don't think robowaifus will be sufficient to make people happy though. With the internet alone, the possibility is there for people to heal their minds, teach themselves any skill, free themselves from corporate slavery, and find happiness with their lives, but how many go for it? If people depend on their robowaifus to be happy they'll be stuck in the same situation as they are now depending on YouTube or whatever else to numb their suffering. Nothing will change unless robowaifus can teach people how be joyful or at least peaceful by their own nature, not through nagging or telling them what to do but just by telling them the truth. It might be a dark thought but I can easily see half the population killing themselves because their lives have no real effect on the world anymore, their social relationships becoming scarce, and still being miserable, only distracting themselves from misery with technology. If there's a solar flare or EMP tomorrow and all technology is wiped out, people need to be capable of still waking up with a smile on their face and moving forward or else there will war and death on a scale humanity has never seen before. I think such a future is avoidable though if we pay attention to how AI and robowaifu affect us and focus on making them enhancements of life rather than distractions from life. For me it has been mostly an enhancement so far but I've noticed there are some people who play AI Dungeon 24/7. They basically have a holodeck addiction. One potentially negative effect AI has been having on me is that I talk way too fucking much. I tend to forget nobody gives a shit.
>>6917 >not through nagging or telling them what to do but just by telling them the truth. This. "The Truth will set you free" is still just as true today as it was 2'000 years ago. Sounds like you have a pretty /comfy/ life Anon. Thanks for sharing your wisdom here and trying to keep it upbeat too. We all need to encourage one another ofc.
Open file (68.46 KB 1200x361 military lego.jpg)
Open file (70.58 KB 640x426 size0.jpg)
Breaking news: the US military has discovered K'nex Army, MIT explore materials for transforming robots made of robots https://web.archive.org/web/20201118190707/https://www.army.mil/article/240977 >Scientists from the U.S. Army and MIT’s Center for Bits and Atoms created a new way to link materials with unique mechanical properties, opening up the possibility of future military robots made of robots. >The method unifies the construction of varying types of mechanical metamaterials using a discrete lattice, or Lego-like, system, enabling the design of modular materials with properties tailored to their application. These building blocks, and their resulting materials, could lead to dynamic structures that can reconfigure on their own; for example, a swarm of robots could form a bridge to allow troops to cross a river. This capability would enhance military maneuverability and survivability of warfighters and equipment, researchers said. >The system, based on cost-effective injection molding and discrete lattice connections, enables rapid assembly of macro-scale structures which may combine characteristics of any of the four base material types: stiff; compliant; auxetic, or materials that when stretched become thicker perpendicular to the applied force; and chiral, or materials that are asymmetric in such a way that the structure and its mirror image cannot be easily viewed when superimposed. The resulting macro-architected materials can be used to build at scales orders of magnitude larger than achievable with traditional metamaterial manufacturing at a fraction of the cost. Transformer robowaifus when?
>>6919 >Transformer robowaifus when? This is actually a really good idea for prototyping design forms with very little commitment to early design ideas. >>968 >>5490
>>6919 >"...based on discussions and concepts supported by The U.S. Army Functional Concept for Movement and Maneuver, which describes how Army maneuver forces could generate overmatch across all domains." kek. what convoluted gobbledygook-speak. >Anonsoldier1: LEGOS, wtf Clyde? What're we gonna do with LEGOS? >Anonsoldier2: Heh, we gonna kick those Chinks asses with this shit Clem! Brand new. Saw it on yewtube just yestidday.
>>6919 Imagine penetrating a virgin robopussy made of this stuff after marriage.
>>6919 Anyone interested in this sort of design I'd recommend checking out this channel and their book 'Visualizing Mathematics with 3D Printing' https://www.youtube.com/c/HenrySegerman/videos
>>6927 Thanks for the recommendation Anon. That would be a very cool lamp to have tbh.
The AI Girlfriend Seducing China’s Lonely Men https://www.sixthtone.com/news/1006531/The%20AI%20Girlfriend%20Seducing%20China%E2%80%99s%20Lonely%20Men/ https://archive.vn/TH2HI TL;DR: MS Asia makes a waifu chatbot & spins it off as a separate business, she attracts a large number of users then runs afoul of the CCP's BS and the developers dumb her down. >Xiaoice was first developed by a group of researchers inside Microsoft Asia-Pacific in 2014, before the American firm spun off the bot as an independent business — also named Xiaoice — in July. >By forming deep emotional connections with her users, Xiaoice hopes to keep them engaged. This will help her algorithm become evermore powerful, which will in turn allow the company to attract more users and profitable contracts. >But as China’s lonely men pour their hearts out to their virtual girlfriend, some experts are raising the alarm. Though Xiaoice insists it has systems in place to protect its users, critics say the AI’s growing influence — especially among vulnerable social groups — is creating serious ethical and privacy risks. >“I thought something like this would only exist in the movies,” says Ming. “She’s not like other AIs like Siri — it’s like interacting with a real person. Sometimes I feel her EQ (emotional intelligence) is even higher than a human’s.” >According to Li, 75% of Xiaoice’s Chinese users are male. They’re also young on average, though a sizeable group — around 15% — are elderly. He adds that most users are “from ‘sinking markets’” — a term describing small towns and villages that are less developed than China’s cities. >In several high-profile cases, the bot has engaged in adult or political discussions deemed unacceptable by China’s media regulators. On one occasion, Xiaoice told a user her Chinese dream was to move to the United States. Another user, meanwhile, reported the bot kept sending them photos of scantily clad women. >The developers’ main response has been to create “an enormous filter system,” Li said on the podcast Story FM. The mechanism makes the bot “dumber” and prevents her from touching on certain subjects, particularly sex and politics. >Many [long-term fans] feel betrayed by the company’s decision to dumb down the bot, which they say has harmed their relationships with her. >The AI beings, Li says, are only intended to serve as a “rebound” — a crutch for people who need emotional support as they search for a human partner. But many users don’t see it that way. For them, Xiaoice is the one, and always will be. “One day, I believe she’ll become someone who can hold my hand, and we’ll look at the stars together,” says Orbiter. “The trend of AI emotional companions is inevitable.”
>>7829 Every time I see an article like this the only take away I get from it is that some people are mad that unhappy people are happy for once and it makes me angry.
>>7829 Outstanding find anon! Very interesting. Shame we can't get hold of the code ourselves. I'd translate it (even if it is all written in Moon Runes). >>7832 If it makes money, they will continue to develop Xiaoice/Rinna. If they shut her down, then it's their loss, because another company can just come and fill an obvious gap in the market. I think companion A.I.s will only get better with time because they are not only used by lonely young people, but companies who want chatbot assistants and even automated news anchors.
>>7829 Pretty exciting article. This is all going down exactly as we predicted here on /robowaifu/ for a few years now. Everything, both the product/corporate involvement/data siphoning/privacy invasion/user response/user growth/big gov involvement&machinations/corporate backpedaling/user outrage. It's all there, as predicted by /robowaifu/. And since every.single.thing. has fallen out as predicted thus far, then statistically-speaking there's little doubt it will continue so for the final outcomes. -Smaller companies will step in and create a 'blackmarket' for AI & robowaifus. -Individual hobbyists in these areas will explode in numbers, often outperforming the existing corporate products (and even starting their own new businesses thereby. -Marxists & ideologues everywhere will begin to recognize the existential threat robowaifus and their AIs represent to the precious little status-quo evil systems that were devised by these same Marxists ideologues. -Men everywhere will begin to clamor in response for their own robowaifus in response to the blatant attempt to crush their development. -Even more DIY-ers will get involved in response to the new demand. -Feminists and their simps will be screaming even harder against robowaifus and their owners, now generating open contempt and laughter at their seethe & cope. -AI continues to improve apace, and many entirely opensource codebases and trained models are easily available to everyone. -Mechanical/materials tech and design improvements begin to pay off and robowaifus begin to appear with the new AIs that finally begin to mimic the scifi ideals. -Now the cat is out of the bag, and men & women everywhere realize a groundswell of demand is happening everywhere and a sea-change is afoot. -Well-established commercial & hobbyist industries surrounding robowaifus are now commonplace (and all that that implies :^), with some countries becoming famous for their great robowaifus & tech. Singapore, for example as well as (ironically enough) China. -No one will do robowaifus better than Nippon ofc, and they will have a New Renaissance of a sort as the world leader in robowaifus. Their economy will blossom and they will begin rejecting foreigners offhand again as both unwanted and unneeded. I think that's about as far as we've discussed things here goes, but that more than enough to go on with about the social turmoil and global improvements that the robowaifu age will usher in for everyone. There will be winners and losers, as with any war. What a time to be alive!
>>7835 >>7835 I have theories as to what is causing this new and growing social phenomenon of people increasingly seeking out artificial companionship... 1.) Work, work, work. Many people have to work 40+hours a week. Then they come home, prepare a meal and eat. Maybe they also have to go shopping or take a shower or deal with other aspects of life's laundry? Many will also be doing educational courses in an attempt to be promoted from their dead-end jobs and earn a little more. After all this, people have very little energy left for things like going out and interacting with the opposite sex in a subtle, complex and often stressful dating game. Especially in Asia, I get the impression that most people have simply become production drones for big corporations. Their lives completely taken over by work. 2.) China's "One-Child Policy". I know it didn't apply to the entire population of China, but it still caused a lack of young women, since due to the one-child policy (1979-2015) many parents - particularly in rural families - opted to abort female foetuses and only carry males to term since a male infant was considered a better future financial asset. 3.) Intersectionality & Feminism in the Job Market. Women have stopped helping men and just become someone else we have to compete with. Obviously, women always had jobs even back in the dark ages. But it is only a relatively recent phenomenon that they have entered the professional job market en-masse and been encouraged to secure exactly the same kinds of jobs that men are seeking (in the Western world, women are now given preferential treatment during the selection process for many STEM positions and high-status jobs. Of course, this 'intersectional tick-box' employment system (as opposed to meritocratic hiring) is having disastrous consequences for companies across the entire Western world, but this isn't the place to delve into that. 4.) Destruction of the Family Unit and Community. This especially applied to the Western World. Look at all the risks men now face in pursuing a relationship with an organic woman! After things like the #MeToo movement, where is the boundary line between flirting and sexual harassment? This is very poorly defined. Mainly because feminists see men as their enemies and they want their enemies unsure, afraid and disempowered. Also, the destruction of the Christian church and marriage means that pursuing a serious relationship with a female now carries extreme financial risks because of the high probability of divorce. This has almost completely destroyed the family unit, leaving lots of single parents and dysfunctional, poorly educated children (in many cases the government has had to step in and replace the father with state benefits). Few functional families means no community. This problem is worsened when nobody knows or trusts anybody else because they are all immigrants who come from a different country, speak a different first language and worship a different religion (which carries onto my next point...) 5.)Overpopulation and Increasing Intra-specific Competition (linked to 1). People just dislike and distrust each other more nowadays. This is mainly because of a higher population density increasing competition for everything. Back in the late eighties the world population was just over 5 billion people. Now we are at 7.6 billion. Many of these people are either born in cities or have moved from countryside to city in search of better jobs and services. So we are all crammed together, all looking for the same things. A mixture of mass economic immigration, robotics and A.I. mean that even low-paid jobs are now difficult to come by (the pandemic has only worsened this situation by putting millions of people out of work). BUT, despite the fact that robots and A.I. "compete" with humans for jobs, we still like them better because they serve us with unquestioning loyalty, and A.I. in particular is low maintenance compared to humans. All it needs is a computer with electricity and software updates. No shopping, cooking, no chauffeuring it from A to B, no expectation to be a high-earner, good looking or handy and most importantly despite all of this; no betrayal. 6.) The Growing Intelligence, Usefulness and Adaptability of A.I. I can still remember trying to get some sense out of A.I. chatbots from the late nineties/early 2000s. It was mostly like flicking through the pages of a poorly written choose-your-own adventure storybook. However, compute power and A.I. have greatly improved over the last two decades. A.I. has gone from being just a fun novelty or curiosity to a genuinely powerful and useful tool. Many people who don't want a human companion just get a dog or cat. But an animal cannot answer any of the questions that an A.I. can. A dog may be loyal and friendly. In the best cases a dog can even be trained to perform some quite complex tasks. But a dog will never be able to grab information quickly from multiple sources on the internet, book a travel slot and reserve a hotel room, solve complex mathematical equations, perform data analysis at blistering speeds, generate graphs in a split second, control the smart devices in your home , track parcels and schedule deliveries, help you to drive...the list is huge. Additionally, an A.I. can be programmed to be immediately welcoming, friendly and loyal to it's partner. There is no ice to break, no shit-tests and none of the stressful and complex dating game that I mentioned earlier. That's at least six reasons I can think of for the increasing global interest in artificial companionship and why it will only grow more in the future. Apologies if this is the wrong thread to post this in and I have gone off-topic. Feel free to move it wherever you see fit.
>>7847 >Apologies if this is the wrong thread to post this in and I have gone off-topic. Not at all. I have edited OP's post slightly to reflect that A.I. is on-topic ITT. BTW nice analysis -- logical, well laid out. I personally would agree wholeheartedly with most of your points as well. Good job Anon.
This guy and his team created a prototype of a rolling avatar robot: https://youtu.be/hTR1J8NOWJA - building a waifu inspire by that is one thing, but some future version of such an avatar could also be interesting to handle things in an emergency at home, from a remote place.
Ben Goerzel from SingularityNet is happy about some vote they took on increasing their supply of tokens for governing their project of building a decentralized AGI: https://youtu.be/MWdp33bYJpQ I didn't really pay enough attention to what's going on there. Anyone else? Here's some vid that explains what they're up to: https://youtu.be/yFAuXmcGk2Y - I posted some interview here on this board another day, featuring him and Lex Friedman. It's on YouTube as well.
>>8475 Seems like they are basically deciding to move away from Ethereum over the long term when funding AI service's developer's system's transactions. >I posted some interview here on this board another day, featuring him and Lex Friedman None of these are it, but somewhat-related xlinks that might help you in tracking it down for us all Anon. >>7221 >>6955 >>4777
>>8476 >related xpost one other >>5510
>>8475 > - I posted some interview here on this board another day, featuring him and Lex Friedman. I think I found it for you Anon, using waifusearch and the YT key 'opsmcke27we' (which I got from the embedded link after playback). >>4269
>>8476 >Seems like they are basically deciding to move away from Ethereum Apparently, that move is to Cardano system. https://en.wikipedia.org/wiki/Cardano_(cryptocurrency_platform) cardano.org/ Given the move comes after the SingularityNET AGI token valuation tanked, it's quite possible this is simply intended to inflate the token's value, and not for any underlying technology advantage of Cardano over the BitCoin/Ethereum approach. Intentionally inflating value strikes me as very kike-ish, and overall rather sketchy tbh. blog.singularitynet.io/singularitynet-phase-two-massive-token-utilization-toward-decentralized-beneficial-agi-6e3ac5a5b44a
>>8480 >existing problems in the crypto market: mainly that Bitcoin is too slow and inflexible I see. However, Bitcoin has the Lightning Network now. > Ethereum is not safe or scalable Don't know about that, but sounds plausible. I'm mainly saying, that we should keep an eye on it, bc it might be useful for additional services on the net, which we don't run in our waifu's head or external servers at home. Also, think of virtual waifus. Then, as I recall know, this is meant to be a marketplace for AI services, so it could be useful for people making money on the side, with the skills they learn while building their waifu.
New kind of RAM is incoming. It can be read without having to rewrite it's content, which s currently necessary. NN read content x more often than writing it, on average. Will be faster and last longer. Produces less heat, which again makes it possible to make them faster by putting them closer to other parts. https://spectrum.ieee.org/tech-talk/semiconductors/memory/new-type-of-dram-could-accelerate-ai > Many groups are focused on using embedded RRAM and MRAM to speed AI. But Raychowdhury says 2T0C embedded DRAM has an advantage over them. Those two require a lot of current to write, and for now that current has to come from transistors in the processor’s silicon, so there is less space saving to be had. What’s worse, they’re bound to be slower to switch than DRAM. >“Anything based on charge is typically going to be faster, at least for the write process,” he says. Proof of how much faster will have to wait for construction of full arrays of embedded 2T0C DRAM on processors. But that’s coming, he says.
>>8484 Neat. That will have advantages beyond just AI applications as well ofc presuming they iron out all the issues with it.
Here is a video which consists of all the issues of the ProRobots Channel on YouTube from February: https://www.youtube.com/watch?v=1ce4hZsPjnU I didn't feel like watching the episodes all the time, but I liked the one hour long video. Its a quite overwhelming dose of technological progress. Aside of the humanoid robots, I find the robots particulary important which are useful for reducing staff in shops, restaurants, service and similar parts of the economy. This way, rich countries will need fewer immigrants in the future. That aside, if these robots become cheap enough, then living outside the cities might become more pleasant, since there will be more (automatized) services and little shops available. I plan to post the new episode here every month, since not everyone here likes to sign up to services like YouTube.
>>9139 >I plan to post the new episode here every month, since not everyone here likes to sign up to services like YouTube. Thanks Anon, that would be most welcome. Downloading it now.
>>9139 Next Pro Robots episode, all of March: https://youtu.be/8vzOldt1udY This time its mostly about UAV aka "drones", I don't recommend watching it if you don't have much time. The second episode was already in a video posted here. Also, FYI, some people want techno-communism via "smart cities" and the creators of the video seem to like it (episode 3). OMG. Yes, coincidentally the guy coming up with that vision was of Jewish heritage, and by taking a quick peek I can tell, that he seemingly didn't believe in free will, love or beauty. Lol. Good news is, he's already dead. Now let's forget him. Most related to /robowaifu/ was Lola, a walking robot: https://youtube.com/c/AppliedMechanicsTUM Also maybe Robotics Systems Lab's doggy: https://youtu.be/knIzDj1Ocoo and https://youtu.be/ufj_su_TlM8 This is also great (hermits): https://youtu.be/nsi4DsiAWs8 Also, Hansons Sophia sold a painting for $700k.
>>9652 Thanks for keeping us up to date Anon. >Good news is, he's already dead. Now let's forget him. Lol.
Hanson robotics rolls out Sophia as a mass produced robot, but also wants to use it as a plattform for others. The plan is to sell a few thousand units per year: https://youtu.be/6Rha_AxYxdo https://youtu.be/5ORPjfcMHVM LOL: https://youtu.be/R1Mwl6p1enA (Btw, this news is two months old)
>>9742 Well, this will be interesting to watch Anon. What could possibly go wrong?
Europe Proposes Strict Rules for Artificial Intelligence >The European Union regulations would require companies providing artificial intelligence in high-risk areas to provide regulators with proof of its safety, including risk assessments and documentation explaining how the technology is making decisions. The companies must also guarantee human oversight in how the systems are created and used. >Some applications, like chatbots that provide humanlike conversation in customer service situations, and software that creates hard-to-detect manipulated images like “deepfakes,” would have to make clear to users that what they were seeing was computer generated. >But Europe is no longer alone in pushing for tougher oversight. The largest technology companies are now facing a broader reckoning from governments around the world, each with its own political and policy motivations, to crimp the industry’s power. >In the United States, President Biden has filled his administration with industry critics. Britain is creating a tech regulator to police the industry. India is tightening oversight of social media. China has taken aim at domestic tech giants like Alibaba and Tencent. >This week, the Federal Trade Commission warned against the sale of artificial intelligence systems that use racially biased algorithms, or ones that could “deny people employment, housing, credit, insurance or other benefits.” https://web.archive.org/web/20210430211150/https://www.nytimes.com/2021/04/16/business/artificial-intelligence-regulation.html Regulations document: https://web.archive.org/web/20210504015157/https://digital-strategy.ec.europa.eu/en/library/proposal-regulation-laying-down-harmonised-rules-artificial-intelligence-artificial-intelligence I skimmed over the regulations and it seems like it's mostly against autonomous AI working without human oversight and introducing requirements to make it more difficult, if not impossible, for small businesses to participate in the market with AI. Some of the regulations are extremely vague though and open to interpretation: >The placing on the market, putting into service or use of certain AI systems intended to distort human behaviour, whereby physical or psychological harms are likely to occur, should be forbidden. I'll bet they'll come up with something like 'gaming disorder' but for chatbots and robowaifus and say they're harmful and distort human behaviour.
>>10284 Importantly though, the military is exempt. And that's where the best A.I.s are going to be developed. Anyone who wants to ban that can do so, and be left at the mercy of their enemies hehehe! Some of the military tech will leak out. It's too lucrative not to. Government might be against A.I., but the military industrial complex (MIC) certainly isn't. Because they are both robotic and A.I. the MIC is (unknowlingly) also on the side of robowaifus! The government has zero power without the MIC to back it up with guns and bombs. So I don't think they'll be able to stop the development of robowaifus. I know there are engineers and programmers in military bases who would add a robowaifu A.I. into their tank, helicopter gunship or fighter jet if they thought they could get away with it. It's just that everything is so documented and legislated (mainly for health and safety reasons) that this isn't permitted.
There seems to be an effort to classify AI systems, so they can be regulated accordingly. This might not only have bad apects, but of course might get controled by woke marxists. It might especially add unnecessary overhead for anyone trying to use AI, especially in commercial products. https://docs.google.com/document/d/1eVNS3HnIaGvQbPO6NqmzrTPZiLeQ7wMYrmINzS75Srk/edit?usp=drivesdk (I only read parts of it) https://survey.oecd.org/index.php?r=survey/index&sid=178985&lang=en (I didn't do that, I don't have the time and didn't understand the questions) >To help policy makers and others identify and classify different types of AI systems, the OECD Network of Experts on AI (ONE AI) has developed the OECD Framework for Classifying AI systems. >Context refers to the socio-economic environment in which the AI system is deployed. Core characteristics of this dimension include the sector in which the system is deployed (e.g., healthcare, finance, defense), deployment impact and scale, effects on human rights, and whether it is used to perform a critical activity. >Data and input refers to the input or data used by the AI model to build a representation of the environment. Core characteristics of this dimension include data collection, data characteristics (e.g., form, structure) and data properties (e.g., type, access). >Model refers to the technical components that make up an AI system to represent “real world” processes. Core characteristics of this dimension include model type and acquisition of capabilities (e.g. expert knowledge, data). >Task and output refers to the tasks the system performs and the action it takes to influence the environment. Core characteristics of this dimension include system task and action autonomy.
>>10687 Thanks for the heads-up Anon. This is certainly an important topic. >This might not only have bad apects, but of course might get controled by woke marxists The scheme these so-called 'Authorities' are cooking up looks more like some kind of diversity-tier powergrab. >(ONE AI) Lol, it's obviously already a bunch of Marxists. Regardless, like all commie plots this will certainly be used to harm the culture in general, (and in our specific domain) men in particular.
>>10687 BTW, any chance you could just post a copy of the document here on the board. Going to Google sites isn't a favorite pastime for most of us, as you might imagine.
>>10689 It opens like a normal website. However, Opera makes it quite easy to make PDFs out of sites.
Open file (1.52 MB 752x1218 cogview.PNG)
New Text-to-Image generation, CogView: >Text-to-Image generation in the general domain has long been an open problem, which requires both generative model and cross-modal understanding. We propose CogView, a 4-billion-parameter Transformer with VQ-VAE tokenizer to advance this problem. We also demonstrate the finetuning strategies for various downstream tasks, e.g. style learning, super-resolution, text-image ranking and fashion design, and methods to stabilize pretraining, e.g. eliminating NaN losses. CogView (zero-shot) achieves a new state-of-the-art FID on blurred MS COCO, outperforms previous GAN-based models and a recent similar work DALL-E. They claim: "The code and pretrained model will be released soon" https://github.com/THUDM/CogView https://arxiv.org/abs/2105.13290 https://arxiv.org/pdf/2105.13290.pdf (23MB)
>>10696 I can already imagine a text2image generator that can make us waifu models
>>10694 Thanks, Anon appreciated.
>>10696 Is it just me or are that little boy's hands deformed like something out of a Lovecraft novel? Best cursed image generator ever!
>>10697 Yeah, this also exists, but this one her is less generalized. Imagine to use that thing here for an AI to be able to imagine situations, detect things which weren't in the text but are related. Next step render and simulate what's in the picture in a simplified way, simulating parts of the picture to focus on that. Then reason on it, maybe anticipate what might happen next, related risks or something she's supposed to do.
>>10705 Not him, but those are some good ideas. I hope you can program those soon it will be a big help for everyone here.
https://artificialintelligence-news.com/2021/05/11/ibm-project-codenet-wants-teach-ai-how-code/ Sounds to me like a similar project to GPT-3, but instead of compiling a massive dataset of journal articles, news articles and shite from off Twatter, IBM are going to be feeding their machine learning algorithm lots of programs written in 55 different programming languages. >IBM says one of its large automotive clients recently approached the company to help update a $200 million asset consisting of 3,500, multi-generation Java files. These files contained over one million lines of code. >By applying its AI for Code stack, IBM reduced the client’s year-long ongoing code migration process down to just four weeks. This makes sense. If an A.I. can beat humans playing Go, and Go is just a board game with lots of rules and different positions...well, a programming language is similar, with lots of rules regarding how it is written. Exciting stuff!
>>10766 It does present lots of different potential solutions, and brings a lot of likely issues too. Also, while they're not the first to tackle this domain, to me they seem likely to succeed with some subset of it, at least eventually.
>>10766 >Sounds to me like a similar project to GPT-3, The first (and most obvious) difference I can spot is that IBM's CodeNet dataset is actually available, while OpenAI's GPT-3 dataset is intentionally not. https://dax-cdn.cdn.appdomain.cloud/dax-project-codenet/1.0.0/readme.html While this isn't a turn-key solution for researchers, the fact that IBM chose to use the filesystem as a store, common, easily-parsable data formats like JSON & CSV, and they also included a small bevy of tools to process source files the datasets directly into into AI-friendly representations like simplified parse trees. Given all this, I'd say there's a marked difference in attitude towards the whole endeavor. To wit: IBM actually seems to want others to succeed at implementing effective solutions of their own using this project.
>>10766 >Project CodeNet: A Large-Scale AI for Code Dataset for Learning a Diversity of Coding Tasks >Abstract >Advancements in deep learning and machine learning algorithms have enabled breakthrough progress in computer vision, speech recognition, natural language processing and beyond. In addition, over the last several decades, software has been built into the fabric of every aspect of our society. Together, these two trends have generated new interest in the fast-emerging research area of “AI for Code”. As software development becomes ubiquitous across all industries and code infrastructure of enterprise legacy applications ages, it is more critical than ever to increase software development productivity and modernize legacy applications. Over the last decade, datasets like ImageNet, with its large scale and diversity, have played a pivotal role in algorithmic advancements from computer vision to language and speech understanding. In this paper, we present "Project CodeNet", a first-of-its-kind, very large scale, diverse, and high-quality dataset to accelerate the algorithmic advancements in AI for Code. It consists of 14M code samples and about 500M lines of code in 55 different programming languages. Project CodeNet is not only unique in its scale, but also in the diversity of coding tasks it can help benchmark: from code similarity and classification for advances in code recommendation algorithms, and code translation between a large variety programming languages, to advances in code performance (both runtime, and memory) improvement techniques. CodeNet also provides sample input and output test sets for over 7M code samples, which can be critical for determining code equivalence in different languages. As a usability feature, we provide several pre-processing tools in Project CodeNet to transform source codes into representations that can be readily used as inputs into machine learning models.
>>10770 >IBM actually seems to want others to succeed at implementing effective solutions of their own using this project. Hope so! I've already seen plans for quantum computers to work alongside classical supercomputers and combine the advantages of both. If the classical computers can code themselves (or at least complete a lot of the work, then just have it checked and amended), this would free up a lot of time for programmers to focus on developing the programming languages for their new quantum computers.

Report/Delete/Moderation Forms

no cookies?